On Thu, Nov 27, 2014 at 1:31 PM, Leonardo M. Rocha <[email protected]>
wrote:

>
> There is a difficulty here, that if it learns continuously, how to
> evaluate the success of the task?
>

As with all tutors, you evaluate their performance by looking at changes in
student performance.   You ask "did the student actually learn?"

A more sophisticated inquiry would ask if your machine can improve the
performance of ALL students, including students who are non-native English
speakers, students who are at risk of failing, students who are doing very
well academically, special education students and so on.

The tutor would really be a planning machine.  It first has to figure out
where the student is.  Then look where we want the student to be and then
find a route from here to there that moves in "right sized" steps.  Then
the machine executes the plan while it continuously evaluates progress and
re-plans as required.

The problem is going to be that to do this the tutor needs an internal
model of the student, that is a hard problem


-- 

Chris Albertson
Redondo Beach, California

Reply via email to